DESCRIPTION: It has become increasingly clear that there are significant weaknesses in peptide and viral protein vaccine approaches in generating potent and sustained anti-human immunodeficiency virus type I (HIV-1) immune responses. Although live, attenuated lentiviral vaccines have demonstrated significant protection, their potential safety has led to substantial concerns. Based on those data, the investigators hypothesize that continually replicating viral vectors, which do not integrate into the host cell's genome, may be critical in generating an effective immune response. Thus, a lowly replicating viral vector would continuously supply the immune system with processed antigens. Systems for generating such a novel vaccine may be obtained utilizing rhabdovirus-based approaches. In previous studies, the investigators have demonstrated that rabies virus (RV) can be used as a highly stable expression vector. These data suggested that RV-based vectors might be an excellent choice as an HIV-1 vaccine. The applicants propose to construct several non-defective and defective RV vectors expressing HIV-1 gpl60 envelopes of laboratory and primary isolates and study anti-HIV-1 antibody and cellular responses in a mouse model. In addition, they will generate a RV vector, which completely lacks any neurotropism. Because HIV-1 gp 160 can substitute for RV glycoprotein, they also propose the construction of recombinant RVs, which will be restricted to CD4+ cells. These RVs mimic the route of infection of HIV-1, which may be necessary to induce a protective immune response against HIV-1. Thus, these focused and complementary specific aims will lead to the development of rabies virus-based vectors, as potential novel vaccine approaches against HIV-1 infection. These approaches could be potentially utilized as prophylactic and/or therapeutic vaccines.